Together, these four engines cover roughly 99% of "find me a lawyer" AI queries in the United States. Each one builds answers from a different signal mix. Each one requires different work to appear in.
The single largest consumer surface. Returns three to five firm names in plain-language answers. Heavily weights citation density and entity recognition.
Citation-first answer engine. Footnotes every claim. Surfaces firms with strong third-party authority and detailed source documents.
Pulls heavily from Google Business Profile, local pack data, and structured review content. Firms with strong local signals win here.
The AI summary above traditional search. Drives a 61% drop in organic clicks when shown. Highest combined volume of any engine.
Visibility on one engine does not guarantee visibility on any other. We routinely see firms ranked in Perplexity (because they have strong cited content) but absent from Gemini (because their Google Business Profile data is weak). Or visible in Google AI Overviews (because their schema is strong) but missing from ChatGPT (because their citation surface across the legal web is thin).
The four engines together represent four different ways the model can fail your firm. Testing all four reveals the specific gap in your signal stack — not just the symptom that "AI isn't showing us." The diagnostic on this site runs the abbreviated cross-engine version. The full audit runs the deeper test.
Google AI Overviews has the highest raw query volume. ChatGPT has the highest "decision query" volume — users entering ChatGPT to choose between options, rather than to research broadly. Perplexity has the most research-mode intent. Gemini's volume comes increasingly from voice and Android assistant queries.
For high-intent "I need to call a lawyer right now" queries, ChatGPT and Google AIO are the two largest sources of decision-stage clients. For research queries that eventually convert ("settlement timelines," "what to expect in mediation"), Perplexity is disproportionately important.
Coarse summary of what each engine over-weights, based on observed retrieval behavior across hundreds of test prompts.
The implication: an AEO build that addresses all four signal mixes simultaneously is more efficient than four single-engine campaigns. The fixes overlap. The work compounds.
Per-engine results. Score from 0 to 16. No email gate.
Run the Diagnostic →